极值BMM模型与POT模型对面板堆石坝安全监控指标拟定的比较研究OACSTPCD
Comparative Study of Extreme BMM Model and POT Model on the Formulation of Safety Monitoring Indicators of Panel Rockfill Dam
研究极值理论中极值BMM模型和POT模型在大坝安全监控指标拟定上的运用,通过引用经济学上的分位数和极端风险(VaR)值的概念,根据对应公式计算出相应数值进行量化比较分析.研究结果表明,极值BMM模型和POT模型在高置信水平下均能反映出数据的极端风险状况;在99%置信水平下两模型VaR值均大于在95%置信水平下的VaR值;在99%置信水平下极值BMM模型计算出的VaR值比POT模型下的VaR值更大,且接近于拟定指标,表明极值BMM模型下VaR值更能反映出数据的极值状态;在指标拟定方面,极值BMM模型较为适合拟定数据范围内的警戒指标,而POT模型指标拟定结果是超越样本数据本身,具有一定的预测性,适合拟定为样本数据的预测指标.
The application of extreme BMM model and POT model in the extreme value theory on the formulation of dam safety monitoring indicators are studied.By referring to the concepts of quantile and extreme risk(VaR)value in economics,the corresponding values are calculated according to the corresponding formula for quantitative comparison and analysis.The results show that both extreme BMM model and POT model can reflect the extreme risk of data at high confidence level.At 99%confidence level,the VaR values of both models are greater than those at 95%confidence level.At 99%confidence level,the VaR value calculated by extreme BMM model is larger than that under POT model,and close to the proposed index,indicating that the VaR value under extreme BMM model can better reflect the extreme state of data.In terms of indicator formulation,the extreme value BMM model is more suitable for the formulation of warning indicators within the data range,while the POT model indicator formulation results are beyond the sample data itself,with certain predictability,suitable for the formulation of prediction indicators of sample data.
孙锡山;李铮;孙亚运
云南农业大学水利学院,云南 昆明 650051南京水利科学研究院,江苏 南京 210018安徽天翰工程有限责任公司,安徽 马鞍山 243000
水利科学
BMM模型POT模型分位数VaR值μ值指标拟定
BMM modelPOT modelquantileVaR valueµ valueindicator formulation
《水力发电》 2024 (007)
106-110,115 / 6
评论